Increased Excretion of Modified Adenine Nucleosides by Children with Adenosine Dearninase Deficiency

Home , Adenine, Deoxyadenosine, Pseudouridine

Pediatr. Res. 16: 362-369 (1982)

ROCHELLE HIRSCHHORN,'"~ HOWARD RATECH, ARYE RUBINSTEIN, PHOTINI PAPAGEORGIOU, HERNANT KESARWALA, ERWIN GELFAND, AND VIVIEN ROEGNER-MANISCALCO Departments of Medicine and Pathology, New York University School of Medicine, New York, New York [R.H., H.R., and V.R.-M.];Department of Pediatrics, Albert Einstein College of Medicine, Bronx, New York [A.R.]; Department of Pediatrics, Rutgers University Medical ~chool,Piscataway, New Jersey [P.P., and H.K.]; and Department of Pediatrics, Hospital for Sick Children, Toronto, Ontario, Canada [E. G.]

Summary tially in, and prevents proliferation of, irnmunocompetent cells, primarily of the T cell class (2, 5, 6, 23, 38, 49, 54). There is also We have identified seven adenine nucleosides in urines of un- in vivo and/or in vitro evidence for alternative mechanisms of treated adenosine deaminase (ADA) deficient patients, four of toxicity, which would operate via depletion of pyrimidine pools, which (adenosine, 2'-deoxyadenosine, 1-methyladenosine and N6- depletion of phosphoribosyl pyrophosphate and increases in cyclic methyladenosine) have been previously identified in urines of AMP or S-adenosyl homocysteine (16, 21, 24, 40, 46, 55). All of normals and/or ADA deficient patients. We confirm that ADA these mechanisms are dependent on accumulation of the substrates deficient patients excrete markedly increased amounts of 2'-deox- of ADA, adenosine and 2'-deoxyadenosine. yadenosine (582 k 363 versus normal of < 0.1 nmoles/mg creati- In addition to adenosine and 2'-deoxyadenosine, several other nine) and increased amounts of adenosine (29.4 & 5.7 versus modified adenine nucleosides occur naturally (17, 19) and are normal of 4.12 & 1.0 nmoles/mg creatinine). substrates for ADA (1, 18, 43, 56). Such naturally occurring We have found three other modified adenine nucleosides previ- modified adenine nucleosides might be expected to also accumu- ously undetected in human urine. These three compounds are 2'- late in ADA deficient patients and possibly contribute to toxicity. 0-methyladenosine, N6, 2'-O-dimethyladenosine and an as yet incompletely characterized modified adenine nucleoside, R-aden- We have, therefore, sought to determine if these modified adenine nucleosides are uniquely present, or present in increased amounts, osine. Only ADA deficient patients excrete detectable amounts of in urine of ADA deficient patients. We have identified seven 2'-0-methyladenosine (2.1 1.1 versus normal of < 0.1 nmoles/ + compounds in urines of ADA deficients, four of which (adenosine, mg creatinine), whereas both normals and ADA deficient children excrete N6, 2'-Odimethyladenosine and R-adenosine. However, 2'-deoxyadenosine, 1-methyladenosine and N"methy1adenosine) have been previously identified in urines of normals and/or ADA ADA deficient patients do excrete increased amounts of R-aden- deficient patients (4,9,27,28,32,36,37,50). We have additionally osine (5.5 1.0 versus normal of 1.4 k 0.4 nmoles/mg creatinine). + found three other substrates of ADA not ~reviouslvdetected in urine. These three compounds newly identified in urine are 2'-0- Speculation methyladenosine, N6,2'-0-dimethyladenosine and an as yet in- Accumulation in ADA deficient patients of two newly detected completely characterized modified adenine nucleoside, R-adeno- modified adenine nucleosides (2'-0-methyladenosine and R-aden- sine. Only ADA deficient patients excrete detectable amounts of osine) or their metabolites could play a role in explaining the 2'-0-methyladenosine, whereas both normal and ADA deficient profound abnormalities of B cell function seen in ADA deficiency children excrete N"2'-0-dimethyladenosine and the as yet uni- but not in purine nucleoside phosphorylase deficiency. This differ- dentified ADA substrate R-adenosine. However, ADA deficient ential involvement of B cell function is not easily explained by patients excrete increased amounts of R-adenosine. accumulation of deoxytrinucleotides, which occurs in both disorders. MATERIALS AND METHODS Materials. Nucleosides were obtained from Sigma Co. (St. Inherited deficiency of the purine salvage enzyme, adenosine Louis, MO) or P&L (Milwaukee, WI). 2'-0-Methylinosine was deaminase (ADA) results in a fatal infantile syndrome of severe generated by incubation of 2'-0-methyladenosine with calf intes- combined immunodeficiency (ADA-SCID) (10, 15, 22, 25, 30). tinal ADA, type I from Sigma, and purified by high pressure Children with ADA deficiency have a profound defect in both liquid chromatography. 5'-Methylthioadenosine was generated by cellular and humoral immunity, although in 10-15% of cases, the acid hydrolysis of ,S-adenosyl methionine (45). 5'-deoxyadenosine humoral defect may initially be less severe (22). Affected children and N~threonino~arbon~ladenosine were the generous gifts of accumulate and/or excrete markedly increasd amounts of the respectively Dr. G. Elion and Dr. G.B. Chheda. ADA substrates, adenosine and 2'-deoxyadenosine, and the phos- Anion exchange chromatography. Urines (stored frozen at phorylated metabolite, deoxy ATP (dATP) (3, 5, 6, 9, 26-28, 32, -70 "C with 0.01% sodium azide) were chromatographed on an 37, 41, 46, 50). anion exchange column (Biorad AG I-X2, Richmond, CA) essen- Several pathophysiologic mechanisms have been proposed tially as described by Kuttesch, et al. (32) except that four fractions whereby accumulation of ADA substrates and their metabolites were collected. Fraction 1, 1-6 ml; fraction 2, 7-3 1 ml; fraction 3, would result in immunodeficiency (5, 12, 15, 16, 21, 24, 40, 52). 32-52 ml and fraction 4, 53-73 ml (eluting after the addition of The largest body of evidence supports the hypothesis that ~ATP, acid). Adenosine and 2'-deoxyadenosine eluted in fraction 2 and an inhibitor of ribonucleotide reductase, accumulates preferen- adenine in fraction 4, as previously described (32, 36). The elution ADENOSINE DEAMINASE DEFICIENCY 363 profiles of several authentic compounds, including modified ade- Peak shifi and acid hydrolysis. Samples and authentic com- nine and guanine nucleosides and bases are shown in Figure I. pounds were incubated with and without calf intestinal ADA High pressure liquid chromarography (HPLC).Urines (fraction- (Sigma Type 1, St. Lfiuis, MO) (20,42) at 0.08, 1.5 and 15 IU/ml ated and unfractionated) were analyzed by HPLC on C 18 pBon- final concentration for 1-2 h at 37OC and then analyzed by HPLC dapak columns (Waters Co., Milford, MA) by methods previously (Fig. 2 and Table 1). For hydrolysis of nucleosides to bases, peaks described (42) except that the gradient went from 0 to 40% of interest were collected by HPLC, concentrated, reconstituted methanol in 60 min. For further identification, peaks of interest with H20 and an aliquot boiled for 30 min in 0.45 N perchloric were collected, dried under nitrogen, reconstituted with water, and acid and neutralized with KOH. The authentic compounds aden- reanalyzed following either treatment with ADA, acid hydrolysis osine, 2'-deoxyadenosine, 2'-0-methyladenosine, N6-methylad- or chromatography on Afigel601 (Biorad, Richmond, CA). UV enosine, I-methyladenosine, N6,N6,-dimethyladenosineand N6- ratios of compounds were determined by simultaneous monitoring isopentenyladenosine were hydrolyzed under the same conditions at the two wavelengths specified. Retention times of relevant and the reaction products analyzed by HPLC. Nz-methylguano- nucleosides and bases are listed in Table 1. sine eluted in the last of the anion exchange fractions (Fig. I) and

0-0 INOSINE r-• N6UETHYL ADENOSINE ( 0-0 OEOXYINOSINE &-d NsH( DIuETnYL ADENOSINE ' W GUANOSINE )-8 2'0 METHYL ADENOSINE r-r GUANINE 0-0 7 METnlL GUlNOSlNE , B--B N~UETHIL GUINOSlNE I - NIM DIMETHYL GUANOS~NEI

4 8 12 16 20 24 28 JL56 60 64 68 72 74 ELUTION VOLUME (ml) Fig. I. Elution of methylated nucleosides from dowex AG I X2. Authentic compounds were diluted and chromatographed on Dowex AG I X2 columns as described in "Materials and Methods". All the modified adenine nucleosides tested (as well as adenosine) eluted in the sewnd fraction except for I-methyladenosine which eluted in the first and sewnd fractions. Isopentenyl adenosine could not be recovered by this method (not shown).

Table 1. Relative retention times of mod$ed adenine nucleosides and the products generared by deamination or acid hydrolysis Retention time Retention time2 Retention time2 Retention time Deamination Retention time Retention time Compound Adenosine product Inosine Hydrolysis product Adenin I-Methyladenosine' Adenosine' inosine adenine 2'-Deoxyadenosine' 2'-deoxyinosine adenine N" threonino~arbon~ladeno- NT' NT sine 5'-Deoxyadenosine 5'-deoxyinosine adenine R-adenosine' R-inosine R-adenine 2'-0-Methyladenosine' 2'-0-methylinosine adenine N6-Methyladenosine' inosine N6-methyladenine 3'-0-Methyladenosine 3'-0-methylinosine adenine ~",2'-0-Dimethyladenosine' 2'-0-methylinosine adenine N"N"-Dimethyladenosine inosine N"N6,-dimethyladenine lsopentenyladenosine inosine "isopenteny ladenineW3 5'-Methylthioadenosine NT NT Adenosine N'-oxide NT NT 2-Methyladenosine NT 2-methyladenine - -- ' Compounds isolated from urine as well as authentic compounds, treated with ADA and also acid hydrolyzed. Retention times of urinary compounds were identical to those of the authentic wmpounds, as determined by coelution. Inosine and adenine have retention time relative to adenosine (1.00) of 0.58 and 0.70. All closely eluting compounds (e.g.,5'-deoxyadenosine, N~- threoninocarbonyladenosine, 2'-0-methyladenosine and R-adenosine) were injected in mixtures to confirm that they migrated separately from each other and their order of elution. Additional compounds not listed but excluded as R-adenosine on the basis of different retention tiemes include 3'- deoxyadenosine and 5'-deoxy, 5'-methylthioadenosine. 3'-deoxyinosine was also excluded as R-inosine. as The hydrolysis product of isopentenyladenosine is referred to "isopentenyl adenine" for brevity. The hydrolysis. . reaction reportedly. gives- rise to several wmpounds by a less direct route. Authentic isopentenyl adenine -mlutes with isopentenyladenosine in this system. ' NT = not tested. HIRSCHHORN ET AL. B A MODIFIED ADENINE NUCLEOSIDES (- ADA) I URINE (- ADA)

In 2'd ln MODIFIED ADEN lNE NUCLEOSIDES 1 lMeln (tADA) URINE 1 2'0 Me In (tADA) 1

LL c

I I I I I I I I I I I I I I I 1 I 5 10 15 20 25 30 35 40 5 10 15 20 25 33 35 40 MINUTES MINUTES Fig. 2A. Reverse phase HPLC (CIS pBONDAPAK) of modified adenine nucleosides. Elution of authentic adenine nucleosides (upper panel) and their deamination products (lowerpanel) monitored at 260 nm. The elution times of additional modified nucleosides not detected in urine are indicated in Table 1. The authentic compounds were individually incubated with varying concentrations of adenosine deaminase and rechromatographed to determine elution times of the respective deamination products (Table I). The tracing ( lowerpanel) demonstrates the results of treatment with adenosine deaminase of a mixture of the nucleosides present in the upper panel. Fig. 2B. HPLC of urine of an ADA deficient child before and after incubation with exogenous adenosine deaminase. The second anion exchange column fraction (see "Materials and Methods" and Fig. I) from urine of an untreated ADA deficient child was concentrated and analyzed by HPLC, monitoring at 260 nm. The arrow indicates a 4-fold increase in sensitivity to a full scale of 0.02. Peaks coeluting with I-methyladenosine (I Me Ar), adenosine (Ar), 2-deoxyadenosine (2' dAr), an unidentified modified adenine nucleoside (R-adenosine or "X), 2'-0-methyladenosine (7-0 Me Ar), N"-methyladenosine (N%e Ar) and N"2'-0-dimethyladenosine (N"'0 Me Ar) (indicated in the upper panel) all disappeared after incubation with adenosine deaminase (lower panel). Peaks appeared after ADA treatment (lower panel) with the retention times of inosine (derived from deamination of adenosine and N"methy1adenosine). 2'-deoxyinosine (from 2'-deoxyadenosine) and I-methylinosine (from I-methyladenosine). (The presumed and authentic I-methyladenosine, N"2'-0-dimethyladenosine and Ntmethyladenosine peaks required approximately 200- and 20-fold higher concentrations of ADA than those used to completely deaminate adenosine or 2'-deoxyadenosine.) 2'-0-methylinosine (2'-0 Me In) (from 2'-0-methyladenosine and Nt2'-0-dimethyladenosine) and X-inosine elute shortly after I-methylinosine (see Table I) and are "buried" in this chromatogram primarily because of the presence of an unidentified interfering peak eluting shortly after I-methylinosine, the high concentrations of I-methylinosine relative to 2'-0- methylinosine and R-inosine as well as the decreased sensitivity of detection of "inosine" as compared with "adenosine" compounds at 260 nm, the wavelength at which this chromatogram was monitored. The chromatogram is only illustrative and the indicated identifications were based on studies where peaks were collected individually, concentrated and rechromatographed with and without incubation with ADA to determine coelution and authentic compounds, UV ratios and the retention times listed in Table I. For studies of 2'-0-methyladenosine and R-adenosine (peak "X"), the two compounds were also separated on Affigel 601 (see text). it was identified in the HPLC chromatogram both by coelution cally and by in vitro tests of immune function. He was clinically with the authentic compound and by "peak shift" using exogenous healthy at the start of the study. The case history for patient 3 has purine nucleoside phosphorylase in the phosphorolytic direction. also been reported (26). She was critically ill for the first 4 months Affigel 601 columns. Affigel 601 (Biorad, Richmond, CA), a of the study. Patient 4 was a 6-month-old child severely ill with boronate gel that selectively binds cis-diol compounds at high pH, multiple infections and classic manifestations of SCID (14). Three was swollen and packed as directed and eluted as described for of the control patients studied were similar to the ADA deficients separation of deoxy ribonucleotides and ribonucleotides (54), in that they were either clinically critically ill or immunodeficient collecting I ml fractions. 2'-Deoxyadenosine and 2'-0-methylad- (controls 1-3). The ages of controls (6 months to 3 years) essen- enosine eluted in the first 10 ml with a peak at fractions 3-8; tially overlapped with those of the ADA deficients at the onset of adenosine and N"methy1adenosine bound and were eluted only study (6 wk to 2% years). after the change to lower pH with a peak in fractions 19-24. Uric acid was determined using uricase (Sigma, Type I), cre- RESULTS atinine with Sigma Kit 555 and protein by the method of Lowry, et al. (33). Analysis of fractionated urine by HPLC and detection of urinary Patients. Patients I and 2 are sibs whose case histories have ADA substrates. We fractionated urines from ADA deficient and been previously reported (44). Patient I was diagnosed prenatally nonADA deficient, immunodeficient and normal children on as ADA deficient. The diagnosis was confirmed both enzymati- anion exchange columns, collected and lyophilised the fraction ADENOSINE DEAMINASE DEFICIENCY 365 which contains adenine nucleosides (fraction 2, see "Materials and creted markedly lower concentrations of N" + 1-methyladenosine, Methods") and analyzed this fraction by HPLC. We found seven primarily because of markedly decreased excretion of l-methylad- compounds in urines of ADA deficient children that absorbed UV enosine (Table 2). [An additional healthy 6-month-old infant also at 260 nm and behaved as if they were substrates for ADA. Thus, excreted 4-fold higher amounts of I-methyladenosine (40 nmoles/ all seven peaks disappeared from the chromatogram after incu- mg creatinine) than did the healthy 3-year-old children, suggesting bation with exogenous ADA (Figure 2B). that I-methyladenosine excretion decreases with age.] Zdentrfication and quantitation of four ADA substrates known to Three additional urinary ADA substrates. We found three addi- be present in urine. Four of the seven peaks coeluted respective1 tional compounds that behaved as if they were substrates for ADA with I-methyladenosine, adenosine, 2'-deoxyadenosine or N 2- and have not been previously described in urines of normals or of methyladenosine and had the same 250/260 and 280/260 UV ADA deficient patients. The first two of these three compounds ratios as the authentic compounds. When each of the peaks was appeared in urine of ADA deficient patients as a double peak collected individually, an aliquot incubated with ADA and re- eluting shortly after 2'-deoxyadenosine (Figure 2B). The later chromatographed, each of the original peaks disappeared and a eluting component of the double peak eluted with authentic 2'-0- new peak appeared that coeluted with and had the same 250/260 methyladenosine and was degraded by ADA at concentrations and 280/260 UV ratios as the expected authentic product of similar to those required to degrade adenosine and authentic 2'- deamination and which represented over 9090 recovery from the 0-methyladenosine. The earlier eluting component of the double original peak (Table 1 and Fig. 2). The first three compounds peak required higher concentrations of ADA, similar to those identified, I-methyladenosine, adenosine and N6-methyladeno- required for conversion of N"methyladenosine to inosine. When sine, have been previously detected in urine of normals (4) whereas this double peak was collected, lyophilised and applied to an 2'-deoxyadenosine has been found previously only in urine of Affigel 601 boronate gel column, the presumed 2'-0-methylad- ADA deficient children (9, 26, 27, 32, 36, 50). enosine eluted in the void volume of the column, consistent with The four ADA deficient children excreted dramatically in- 2'-0-methylation of the ribose moiety and loss of cis-diol groups. creased amounts of 2'-deoxyadenosine, ranging from 220-1090 The other component bound to Affigel 601 columns and was nmoles/mg creatinine (Table 2), similar to previously reported eluted by lowerkg the pH. We were, thkrefore, able to completely values. NonADA deficient children, whether immunodeficient or separate the two compounds by chromatography on Amgel 601 healthy, excreted undetectable (<0.01 nmoles/mg creatinine) for further characterization. amounts of 2'-deoxyadenosine. Excretion of adenosine was also Further characterization and quantitation of 2'-0-methyladeno- increased, although less markedly, with an average excretion of sine in urine of ADA deficient children. In addition to having 29.4 + 5.7 compared to excretion of 4.1 f 1.0 nmoles/mg creati- identical retention times (as determined by coelution) and lacking nine by the nonADA deficient children. Quantitation of I-meth- vicinal hydroxyl groups, the presumed 2'-0-methyl compound yladenosine or N6-methyladenosine excretion was complicated by shared other properties with authentic 2'-0-methyladenosine. The the fact that the two compounds can interconvert in solution (19). presumed 2'-0-methyladenosine and the compounds generated Combined 1-methyladenosine + N"methy1adenosine excretion by deamination (2'-0-methylinosine) and acid hydrolysis (ade- by the ADA deficient children was only slightly increased (ap- nine) coeluted with and had UV spectral ratios like those of proximately 2-fold) when compared with excretion by the three authentic 2'-0-methyladenosine and the compounds generated by younger children (102 f 24 versus49 + 3.6 nmoles/mg creatinine). deamination and acid hydrolysis of authentic 2'-0-methyladeno- 1-Methyladenosine excretion by both ADA deficient and sine (Tables 1 and 3). nonADA deficient young children was markedly greater than N6- We could not detect any 2'-0-methyladenosine in urines of methyladenosine excretion. The two older, healthy children ex- normal or nonADA deficient, immunodeficient children. We

Table 2. Excretion of adenine nucleosides by ADA deficient and nonA DA deficient children (adenosine, 2'-deoxyadenosine, I-methyl + N6-methvladenosine. 2'-0-methvladenosine and R-adenosine) - -- Compounds'

Subiects Aze dAr ADA deficient I L 3 4 Average + S.D. NonADA deficient2 I L 3 4 5 Average + S.D. ' nmoles/mg creatinine: Ar, adenosine; dAr, 2'-deoxyadenosine; I Me Ar + N%e Ar, I-methyladenosine + NG-methyladenosine;2'0 Me Ar, 2'- methyladenosine and R-Ar, R-adenosine or "Peak X.Compounds were quantitated by determining area as measured by height and width at half peak height, compared with areas of authentic compounds. For R-adenosine and R-inosine, the area/pmole obtained for adenosine and inosine was used. Quantitation of 2'0 Me Ar and R-Ar was performed by using several C18 pBondapak Columns, which either optimally separated the substrates (2'-0- methyladenosine and R-adenosine) or the deamination products (2'0-methylinosine and R-inosine) or both on the same column. Values obtained by the two separate methods of quantitation agreed within 10%.We could not ascertain why the commercially packed columns differed in resolution. We made only limited attempts to modify the conditions and were unsuccessful in obtaining resolution on all columns. Controls: Patients I and 2 were immunodeficient, nonADA deficient infants. Patient 2 was severely ill. Patient 3 was a critically ill, nonimmuno- deficient child. Patients 4 and 5 were healthy, normal children. "eventy days following last transfusion. 366 HIRSCHHORN ET AL.

Table 3. UV ratios of 2'-0-methyladenosine, R-adenosine and the both from normal and ADA deficient urine, was a relatively poor compounds produced by acid hydrolysis or deamination substrate for ADA, requiring concentrations of ADA similar to UV ratios those needed to deaminate N~-methyladenosine. We quantitated excretion of R-adenosine both by measuring 250/260 * 280/260 * the area of the peak, using columns which best separated the Compounds' S.D. S.D. nucleoside from 2'-0-methyladenosine, and by measuring the area of the deamination product, using columns which best separated Adenine nucleosides the deamination vroduct from 2'-0-methylinosine. The two Ar authentic methods were in good agreement, assuming relative extinction 2'0 MeAr authentic coefficients as for adenosine and inosine. The four ADA deficient 2'0 MeAr urinary children excreted between 4.6 and 6.9 nmoles R-adenosine/mg R-Ar urinary creatinine, compared to normal excretion of 1.4 k 0.4 (Table 2).- Bases (products of acid hydrolysis) N"2'-0-Methyladenosine. The third, previously undescribed Ade authentic ADA sensitive urinary compound coeluted with N6,2'-O-dimeth- Ade from authentic 2'0 MeAR yladenosine. When this peak (isolated from both normal and Ade from urinary 2'0 MeAr ADA deficient urine) was incubated with ADA, a peak was R-Ade from urinary R-Ar generated that coeluted with and had the same 250/260 UV ratios Inosine nucleosides iproducts of deamina- as 2'-0-methylinosine, the deamination product of N6,2'-0-di- tion) methyladenosine. This compound was identified in urine of both In authentic 1.61 * 0.04 0.25 * 0.04 normals and ADA deficient patients. The presumed N6,2'-0- 2'0 Meln from authentic 2'0 MeAr 1.54 * 0.04 0.26 * 0.01 dimethyladenosine peak comigrated with at least one other 2'0 MeIn from urinary 2'0 MeAr 1.72 * 0.23 0.24 nonADA sensitive compound and the peak has not as yet been R-In from urinary R-Ar 1.36 -C 0.20 0.28 * 0.04 ' Ar, adenosine; 2'0 MeAr, 2'-0-methyladenosine; Ade, adenine; In, inosine.

quantitated 2'-0-methyladenosine in urines of ADA deficient patients both by measuring the 2'-0-methyladenosine peak and by isolating the peak and measuring the deamination product. Both measurements were in good agreement. 2'-0-Methyladeno- sine excretion ranged between 0.7-2.2 nmoles/mg creatinine in three untreated ADA deficient natients (Table 2). A fourth ADA deficient child excreted 3.5 nmoles/mg creatinine 70 days after a prior erythrocyte transfusion. Urine from three infants with var- ious other immunodeficiency disorders and from two normal children were also analyzed. We could not detect 2'-0-methyladenosine excretion by any of these children with a lower limit of PEAK "Xu + 2'0 Me AI sensitivity ranging from <0.05 to metabolism. Con- Post Tx 3 mo 19.5 14.8 15.0" 4.6" sistent with the hypothesis that the increased excretion of all four Post Tx 24 mo 12.1 9.4 22.9 4.1 f 1.0 adenine nucleosides results from deficiency of ADA, we have also 2'-0-methyladenosine found that the excretion of R-adenosine and 2'-0-methyladeno- Pre Tx 2.1 39 2.2 0.7 sine as well as of adenosine and 2'-deoxyadenosine by ADA Post Tx I mo 2.0 2.3 0.6 0.1'" deficient patients is diminished following "enzyme replacement Post Tx 3 mo 3.3 2.4 0.4 0.2" therapy" with partial erythrocyte exchange transfusions (36,41). Post Tx 24 mo 1.0 0.9 0.2 < 0.1 R-adenosine Modified adenine nucleosides, other than 2'-deoxyadenosine, are constituents of RNA in mammals and other eukaryotes (34, Pre Tx 5.2 5.33 6.9 4.6 39,47, 5 1,53) and therefore their excretion should reflect turnover Post Tx I mo 2.1 2.4 1.9 1.0'" of RNA. Modification of the adenine nucleoside moiety appears Post Tx 3 mo 4.0 1.7 1.7 1.2Ih to occur only post-transcriptionally, after incorporation &to-RNA. Post Tx 24 mo 1.7 1.8 1.5 1.4 f 0.4 It does not appear to involve direct incomration of modified ' nmoles/mg creatinine. Values are for urines obtained 13 days follow- adenine compb;nds into RNA. The extent and type of modifica- ing transfusions except where indicated. " One day post transfusion. tion varies with the different classes of RNA. For example, N"2'- Seven days post transfusion. Patients 1-4 are the same as in Table 2. 0-dimethyl compounds occur in the cap sturcture of mRNA and Patients 1-3 were partially exchange transfused (10-15 ml/kg) every 2-3 in low molecular weight nuclear RNA; 2'-0-methyladenosine wk for the first year and approximately every 5 wk during the second year. occurs in ribosomal RNA as well as in the cap structure of mRNA Patient 4 was totally exchange transfused as well as partially trans- and I-methyladenosine is a feature of tRNA (7, 13, 3 1, 34, 35, 39, fused. 47,48,5 1,53). Whether modified adenine nucleosides in urine are Seventy days following last transfusion. derived from turnover of endogenous RNA, of dietary purines or 368 HIRSCHHORN ET AL. of RNA of enteric bacteria cannot be defmitely determined from REFERENCES AND NOTES the studies reported here but would require "tracer" studies. The I. Agarwal. R. P .• Sagar. S. M. and Parks, R. E. Jr.: Adenosine deaminase from source of increased adenine nucleosides excreted by ADA defi human erythrocytes. Biochem. Pharmacol., 24: 693 ( 1975). cient patients is irrelevant in relation to the possible pathologic 2. Carson, D. A., Kaye, J., and Seegmiller, J. E.: Lymphospecific toxicity in effects of accumulation of such compounds, although the source adenosine deaminase deficiency and purine nucleoside phosphorylase defi may have significance in relation to possible therapeutic modu ciency: possible role of nucleoside kinases. Proc. Nat!. Acad. Sci., 74: 5677 (1977). lation. 3. Chen, S.-H., Ochs, M.D .. Scott, C. R., Giblet!, E. R., and Tingle, A. J.: Adenosine In normal patients, urinary excretion of modified adenine nu deaminase deficiency. Disappearance of adenine deoxynucleotides from a cleosides will, in theory, depend on several factors including the patient's erythrocytes after successful marrow transplantation. J. Clin. Invest., relative abundance of each of the compounds in RNA, the tum 62: 1386 (1978). 4. Chheda, B. G.: Purine, pyrimidine, pyridine, and imidazole derivates excreted in over rates of the RNA in which they occur, the clearance by the human urine. In: G. D. Fasman, Ed.: Handbook of Biochemistry and Molec kidneys of each compound, as well as the relative ability of ADA ular Biology. Vol. I, 3rd Edition. pp. 251-270 (CRC Press, Cleveland, OH, to remove the group attached at the C6 position of each of the 1975). compounds. N 6-methyladenosine, 1-methyladenosine, N6,2'-0-di 5. Cohen, A. R., Hirschhorn, R., Horowitz, S. D., Rubinstein, A., Polmar, S. H., Hong, R., and Martin, D. W. Jr.: Deoxyadenosine triphosphate as a potentially methyladenosine and R-adenosine are poor substrates for calf, toxic metabolite in adenosine deaminase deficiency. Proc. Nat!. Acad. Sci., 75: avian and/or human ADA (1, 8, 18, 43, 56 and unpublished 472 (1978). observations). It is therefore not surprising that these latter four 6. Coleman, M. S., Donofrio, J., Hutton, J. J., Hahn, L., Daoud, A., Lampkin, B., compounds are found unaltered in urine of ADA normal humans. and Dyminski, J.: Identification and quantitation of adenine deoxynucleotides in erythrocytes of a patient with adenosine deaminase deficiency and severe The finding that ADA deficient children excrete increased combined immunodeficiency. J. Bioi. Chern., 253: 1619 (1978). amounts of R-adenosine and possibly 1-methyladenosine, both 7. Crain, P. F., Choi, Y. C., Busch, H., and McCloskey, J. A.: Characterization of relatively poor substrates for calf ADA, suggests that deamination N',2'-0-dimethyladenosine from nuclear RNA of Novikoff Hepatoma. Nucl. by adenosine deaminase may play some role in the metabolism Acids. Res., 5: 771 ( 1978). 8. Daddona, P. E. and Kelley, W. N.: Human adenosine deaminase-purification even of compounds that are relatively poor substrates for the and subunit structure. J. Bioi. Chern., 252: 110 (1977). enzyme. Adenosine, 2' -deoxyadenosine and 2' -0-methyladeno 9. Donofrio, J., Coleman, M. S., Hutton, J. J., Daoud, A., Lampkin, B., and sine are excellent substrates for human ADA as they are for calf Dyminski, J .: Overproduction of adenine deoxynucleosides and deoxynucleo ADA (1, 8, 56 and unpublished observations). Consistent with tides in adenosine deaminase deficiency with severe combined immunodefi ciency disease. J. Clin. Invest., 62: 884 (1978). their rapid deamination by human ADA, 2' -0-methyladenosine 10. Hirschhorn, R.: Clinical delineation of adenosine deaminase deficiency. In: and 2' -deoxyadenosine have not been detected in normal human Enzyme Defects and Immune Dysfunction (Ciba Foundation Series 68). pp. urine but are found in urines of ADA deficient children. Adeno 35-54 (Excepta Medical, Amsterdam, 1979). sine, an excellent substrate for ADA, is found in urine of normals, II. Fain, J. N. and Malbon, C. C.: Regulation of adenylate cyclase by adenosine. although in lower concentrations than in urine of ADA deficients. Molec. Cell Biochem., 25: 143 (1979). 12. Fox, I. H., Kaminska, J., Wortmann, R. L., and Halsey, D. L.: Deoxyadenosine The presence of adenosine in normal urine may reflect the relative inhibits DNA synthesis in cultured human fibroblasts. J. Lab. Clin. Med., 94: abundance of the adenine-ribose moiety, which is present not only 52 (1979). in RNA but exists in high concentration as free adenine nucleo 13. Gauss, D. H., Gruter, F., and Sprinzl, M.: Compilation oftRNA sequences. In: tides and in additional compounds such as S-adenosylmethionine. P. R. Schimmel, D. Soli, J. N. Abelson, Ed.: Transfer RNA: Structure, Properties and Recognition. pp. 520-535 (Cold Spring Harbor, New York, Alternatively, urinary adenosine may be primarily derived directly NY, 1979). from the kidney. 14. Gelfand, E. Personal communication. The possible significance and role of these modified adenine IS. Giblett, E. R., Anderson, J. E., Cohen, F .. Pollara, B., and Meuwissen, H. J.: Adenosine deaminase deficiency in two patients with severely impaired cellular nucleosides in the pathophysiology of ADA deficiency is, at the immunity. Lancet, II: 1067 (1972). moment, speculative. The concentrations of 2' -0-methyladenosine 16. Green, H. and Chan, T.-S.: Pyrimidine starvation induced by adenosine in and R-adenosine we have measured in urine are insignificant fibroblasts and lymphoid cells: role of adenosine deaminase. Science, 182: 836 compared to the massive concentrations of deoxyadenosine; how (1973). ever, concentrations in urine may not reflect concentrations of 17. Hall, R. H.: Modified nucleosides: physical properties, chemical synthesis and natural occurrence. In: The Modified Nucleosides in Nucleic Acids. pp. 17- compounds in plasma or of the compounds and/or their metabo 185 (Columbia University, Press, New York, NY, 1971). 6 2 lites in cells. Indeed, the relative concentration of deoxyadenosine 18. Hall, R. H., Alam, N. S., Mclennan, B. D., Terrine, C., and Guern, J.: N -(A - and adenosine in plasma are essentially equal and do not reflect isopentenyl) adenosine: its conversion to inosine, catalyzed by adneosine amino the massive accumulation of dATP in cells nor of deoxyadenosine hydrolases from chicken bone marrow and calf intestinal mucosa. Can. J. Biochem., 49: 623 ( 1971 ). in urine. Therefore, further investigation of concentrations of the 19. Dunn, D. B., and Hall, R. H.: Natural occurrence of the modified nucleosides. compounds described here and/or their metabolites in plasma and In: G. D. Fasman: Handbook of Biochemistry and Molecular Biology. Vol. I, cells would seem warranted. 3rd Edition. pp. 65-215 (CRC Press, Cleveland, OH, 1975). There are several possible sites at which these compounds or 20. Hartwick, R., Jeffries, A., Krstulovic, A .. and Brown, P. R.: An optimized assay for adenosine using reverse phase high pressure liquid chromatography. J. their phosphorylated derivatives could interfere with normal cel Chroma!. Sci., 16: 427 ( 1978). lular metabolism. These include binding of mRNA to ribosomes, 21. Hershfield, M.S., Kredich, N. M., Ownby, D. R., Ownby. D. R., Ownby, H., inhibition of S adenosyl homocystine hydrolase and thus S ad and Buckley, R.: In vivo inactivation of erythrocyte S-adenosyl homocysteine enosyl methionine dependent methylation or alteration of cyclic hydrolase in adenosine deaminase-deficiency patients. J. Clin. Invest., 63: 807 AMP concentrations via "purinergic" cell receptors (II, 29, 30, (1979). 22. Hirschhorn, R.: Defects of purine metabolism in immunodeficiency diseases. In: 47). We have examined the effect of several of these compounds R. S. Schwartz, Ed.: Progress in Clinical Immunology. Vol. III, pp. 67-83 on the ability of lymphocytes rendered artificially ADA deficient (Grune & Stratton, New York, NY, 1977). to respond to phytohemagglutinin, a response which primarily 23. Hirschhorn, R., Bajaj, S., Borowsky. W., Kowalski, A., Hong, R., Rubinstein, A., reflects function of fully differentiated T cells. We have found that and Papageorgiou, P.: Differential inhibition of adenosine deaminase deficient peripheral blood lymphocytes and lymphoid line cells by deoxyadenosine and 2'-0-methyladenosine is as potent in vitro as adenosine in inhibit adenosine. Cell Immunol., 42:418 (1979). ing lymphocyte response to mitogens but markedly less potent 24. Hirschhorn, R., Grossman, J., and Weissman, G.: Effect of cyclic 3',5'-adenosine than deoxyadenosine. It remains to be determined if methylated monophosphate and theophylline on lymphocyte transformation. Proc. Soc. adenine nucleosides have significant effects on in vitro models of Exp. Bioi. Med., 133: 1361 (1970). 25. Hirschhorn, R. and Martin, D. W. Jr.: Enzyme defects in immunodeficiency differentiation of immune function and particularly of B cell diseases. In: P. Miescher, H. Muller-Eberhard, Ed.: Springer Seminars in function. Since methylated adenine nucleosides would not be Immunopathology. pp. 299-321 (Springer-Verlag, Heidelberg, 1978). expected to accumulate in purine nucleoside phosphorylase defi 26. Hirschhorn, R., Papageorgiou, P. S., Kesarwala, H. H., and Taft, L. T.: Amel ciency such differences might explain the profound impairment of ioration of neurologic abnormalities after "enzyme replacement" in adenosine deaminase eficiency. N. Engl. J. Med., 303: 377 (1980). B cell function observed in ADA deficiency and not in purine 27. Hirschhorn, R., Papageorgiou, P., Rubinstein, A .. and Rosen, F. S.: Transfusion nucleoside phosphorylase deficiency. vs bone marrow transplantation in combined immunodeficiency. Clin. Res., ADENOSINE DEAMINASE DEFICIENCY 369

27: 507A (1979). and possible homology of isozymes of adenosine deaminase in aves and 28. Hirschhorn, R., Roegner, V., Rubinstein. A,, and Papageorgiou. P.: Plasma humans. Enzyme. 26: 74 (1981). deoxyadenosine, adenosine, and erythrocyte deoxy-ATP are elevated at birth 44. Rubinstein, A., Hirschhorn, R., Sicklick. M., and Murphy, R. A.: In vivo and in in an adenosine deaminase-deficient child. J. Clin. Invest.. 65: 768 (1980). vitro effects of thymosin and adenosine deaminase on adenosine-deaminase- 29. Hoffman, J. L.: Biosynthesis of S-Nhethyl adenosine homocysteine. J. Biol. deficient lymphocytes. New Engl. J. Med., 300: 387 (1979). Chem., 253: 2905 (1978). 45. Schlenk. F.. Zydek-Cwick. C. R., and Hutson. N.: Enzymatic deamination of 30. Hirschhorn. R.: Incidence and prenatal detection of adenosine deaminase defi- adenosine sulfur compounds. Arch. Biochem. Biophys., 42: 144 (1971). cient and purine nucleoside phosphorylase deficiency. In: B. Pollara. R. J. 46. Schmalstieg, F. C.. Nelson, J. A,. Mills. G. C., Monahan, T. M., Goldman. A. S.. Pickering. H. G. Meuwissen. I. Porter, Eds.: Inborn Errors of Specific Immu- and Goldblum, R. M.: Increased purine nucleotides in adenosine deaminase- nity. pp. 5-15 (Academic Press, New York. NY, 1979). deficient lymphocytes. J. Pediatr.. 91: 48 (1977). 31. Kaehler, M., Coward, J., and Rouman, F.: In vivo inhibition of Novikoff 47. Shatkin, A. J.: Capping of eucaryotic mRNA's. Cell. 9: 645 (1976). cytoplasmic messenger RNA methylation by S-tubercidinyl homocysteine. 48. Shibata. H.. Ro-Choi. T. S., Reddy. R., Choi, Y. C., Henning, D.. and Busch. H.: Biochem., 16: 5770 (1977). The primary nucleoside sequence of nuclear U 2 ribonucleic acid. J. Biol. 32. Kuttesch. J. F., Schmalstieg, F. C.. and Nelson. J. A,: Analysis of adenosine and Chem.. 250: 3909 (1975). other adenine compounds in patients with immunodeficiency diseases. J. Liq. 49. Simmonds, H. A,. Panayi, G. S.. and Corrigal, V.: The role of purine metabolism Chromat.. 1: 97 (1978). in immunodeficiency: adenosine deaminase activity and deoxyadenosine ca- 33. Lowry. 0. H., Rosebrough, N. J.. Farr. A. L., and Randall, R. J.: Protein tabolism. Lancet. 1: 60 (1978). measurement with the Folin phenol reagent. J. Biol. Chem.. 193: 265 (1951). 50. Simmonds, H. A.. Sahota. A,, Potter. C. F.. and Cameron, J. S.: Purine metabo- 34. Maden. B. E. H.: Eukaryotic ribosomal RNA methylation: summary and recent lism and immunodeficiency: urinary purine excretion as a diagnostic screening developments. In: E. Usdin, R. T. Borchardt, C. R. Creveling, Eds: Trans- test in adenosine deaminase and purine nucleoside phosphorylase deficiency. methylation. pp. 381-398 (Elsevier/North Holland, New York. NY. 1978). Clin. Sci. Molec. Med., 54: 579 (1978). 35. Maden. B. E. H. and Salim, M.: The methylate9 nucleotide sequences in HeLa 51. Soll. D.: Enzymatic modifications of transfer RNA. Science. 173; 293 (1971). cell ribosomal RNA and its precursors. J. Mol. Biol.. 88: 133 (1974). 52. Uberti, J.. Lightbody, J.. and Johnson. R.: The effect of nucleosides and deoxy- 36. Mills. G. C.. Goldblum, R. M.. Newkirk. K. E., and Schmalstieg, F. C.: Urinary coformycin on adenosine and deoxyadenosine inhibition of human lymphocyte excretion of purines, purine nucleosides. and pseudouridine in adenosine activation. J. Immunol.. 123: 189 (1979). deaminase deficiency. Bichem. Med., 20: 180 (1978). 53. Weinberg, R. A. and Penman, S.: Small molecular weight amono disperse nuclear 37. Mills. G. C.. Schmalstieg. F. C., Trimmer, K. B., Goldman, A. S., and Goldblum. RNA. J. Mol. Biol.. 38: 298 (1968). R. M.: Purine metabolism in adenosine deaminase deficiency. Proc. Natl. 54. Wilson. J. M.. Mitchell. B. S.. Daddona, P. E.. and Kelley. W. N.: Purinogenic Acad. Sci., 73: 2867 (1976). immunodeficiency disease: differential effects of deoxyadenosine and deoxy- 38. Mitchell, B. S., Mejias. E.. Daddona. P. E.. and Kelley, W. N.: Purinogenic guanosine on DNA synthesis in human T lymphoblasu. J. Clin. Invest., 64: immunodeficiency diseases: selective toxicity of deoxyribonucleosides for T 1475 (1979). cells. Proc. Natl. Acad. Sci.. 75: 501 1 (1978). 55. Wolberg. G., Zimmerman. T. P.. Hemstra, K.. Winston, M. and Chu. L.-C.: 39. Nishimura. S.: Minor components in transfer RNA. In: J. N. Davidson, W. E. Adenosine inhibition of lymphocyte-mediated cytolysis: possible role of cyclic Cohen. Eds.: Progress in Nucleic Acid Research and Molecular Biology. Vol. adenosine monophosphate. Science. 187: 957 (1975). 12. pp. 49-85 (Academic Press, New York. NY. 1972). 56. Zielke, C. L. and Suelter, C. H.: Purine, purine nucleoside and purine nucleotide 40. Planet. G. and Fox, I. H.: Inhibition of phosphoribosyl pyrophosphate synthesis amino hydrolases. In: P. D. Boyer. Ed.: The Enzymes. Vol. 4, Third Ed. pp. by purine nucleosides in human erythrocytes. J. Biol. Chem., 251: 5839 (1976). 54-64 (Academic Press, New York. NY, 1971). 41. Polmar. S. H.. Stern. R. C.. Schwartt. A. L.. Wetzler. E. M.. Chase. P. A,, and 57. The authors thank Dr. J. Wilner for help in obtaining urines. Hirschhorn, R.: Enzyme replacement therapy for adenosine deaminase defi- 58. Requests for reprints should be addressed to: Dr. Rochelle Hirschhorn. Depan- ciency and severe combined immunodeficiency. New Engl. J. Med., 295: 1337 ment of Medicine and Pathology. New York University School of Medicine. ( 1976). 550 First Avenue, New York. NY 10016. 42. Ratech, H. and Hirschhorn. R.: Identification of adenosine and eight modified 59. This research was supported by NIH A1 10343 and National Foundation #64. adenine nucleosides using reversed-phase high pressure liquid chromatography Howard Ratech was supported by National Cancer Institute Training Grant and enzymatic peak shift with adenosine deaminase. I. Chromat. Biomed. #5 T32 CA 09161. Applicat.. 183: 499 (1980). 60. Received for publication May 13, 1981. 43. Ratech, H.. Thorbecke. G. J., Meredith, G.. and Hirschhorn. R.: Comparison 61. Accepted for publication October 2. 1981.